Everyone’s trading this week like the AI boom just cracked.

On October 22nd, Lam Research flagged a supply problem, and chip stocks lost roughly $200 billion in a single session.

AMD got dragged down 8% in that selloff, right alongside names with genuine demand problems.

Here’s the detail almost nobody repeated: Lam’s own CFO said the shortfall was equipment shortages, not weakening orders.

Supply, not demand.

AMD’s stock didn’t get that memo. It’s still sitting below every major moving average, more than a hundred dollars off its September high.

Meanwhile, three analysts reiterated price targets of $600 to $675 in that same stretch, like nothing happened at all.

So today I’m putting this one on trial: is the Street right that AMD’s business never changed, or is the chart right that something underneath it actually did?

One of them is wrong, and November 3rd is going to tell us which.

Missed Monday’s edition on Tesla? 🤔 Tesla’s Spending Cut Might Not Be Real

Anthropic has publicly filed for an IPO, giving investors their first detailed look at the economics behind Claude. Revenue has surged, but the filing also highlights enormous spending requirements, security risks, and the cost of staying competitive at the frontier of AI. The bigger takeaway is that the IPO could become a major test of whether public markets are willing to fund AI companies at trillion-dollar valuations despite massive capital needs.

Consumer staples have rebounded sharply in 2026, with Coca-Cola, Monster Beverage, and Keurig Dr Pepper each up more than 20% this year. Coca-Cola has been the standout, rising more than 30% while continuing to grow both pricing and volumes. In plain English, investors appear to be rotating toward durable cash flows and defensive growth, and beverages are emerging as one of the strongest pockets of the sector.

Bristol Myers’ experimental lung-cancer drug iza-bren delivered median progression-free survival of 11.1 months versus 5.7 months for chemotherapy in a Phase 3 trial. That is encouraging for a company trying to replace revenue from aging blockbuster drugs, but investors stayed cautious because the study was conducted in China and broader regulatory validation is still needed. The takeaway is that the drug could become meaningful, but it is not yet a clean blockbuster story.

Invest in single-asset private real estate opportunities alongside a $12B AUM owner/operator. Lightstone invests 20% or more of the equity in every offering, with a $100K minimum investment for accredited investors.

Explore Opportunities

^{Any message that includes model/proforma/targeted figures need a disclosure at the end such as: "This communication is for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to buy any securities. Any such offer or solicitation will be made exclusively through the definitive offering documents. All investments involve risk of loss, including the potential loss of principal. Past performance is not a guarantee of future results. Any targeted returns or projections are forward-looking statements, are based on current assumptions, and are not guarantees of future performance. Actual results may differ materially.}

AMD closed around $518 on Monday, roughly 17% below the all-time high it set a month ago.

In that same stretch, three separate analysts reiterated or raised their price targets into the $600 to $675 range.

That’s not a small gap. That’s the Street pricing in 20% to 30% upside on a stock whose own chart is still pointing the other way.

Here’s the question I’m putting on trial today: is the Street right that AMD’s fundamentals never changed, or is the chart right that something underneath the story is actually breaking?

AMD guided Q3 revenue to about $13 billion, give or take $300 million, with gross margin near 56%.

Consensus sits right there, at $13.06 billion, up 41% year over year, and GAAP EPS of $1.42, more than double last year’s number.

The Data Center segment grew 107% year over year last quarter. Management is calling for “strong double-digit” sequential growth again in Q3, and more than doubling that segment again in 2027.

Server CPUs are guided for over 80% year-over-year growth in the back half of this year, and more than 70% again next year.

None of that reads like a company with a demand problem.

Layer on the Helios rack-scale platform, now ramping toward anchor customers, OpenAI, Meta and Anthropic, each deploying at multi-gigawatt scale.

Three analysts backed this with fresh calls this month. TD Cowen’s Joshua Buchalter reiterated Buy at $675. Susquehanna’s Christopher Rolland raised his target from $550 to $620. Wedbush’s Matt Bryson kept Outperform at $600. And these calls came after the Lam Research selloff, not before it. Nobody downgraded.

That’s the bull case the market believed a month ago, and on the numbers alone, nothing in it has actually changed.

Here’s where the “it was never about AMD” defense gets complicated.

Lam Research’s CFO, Doug Bettinger, said the shortfall was component shortages, with some shipments slipping from the December quarter into March.

Supply, not demand. That much is accurate.

But AMD doesn’t make its own chips. It depends on a supply chain that runs straight through the exact equipment makers now warning about shortages.

If Lam, Applied Materials and ASML can’t ship tooling on schedule, that eventually shows up as capacity constraints further down the chain, the same chain AMD needs to ramp Helios through 2027.

Because supply constraints anywhere in that chain eventually become AMD’s constraint too.

The market didn’t misread the warning. It read through it, straight to the part that actually matters for AMD’s own 2027 math.

That’s the uncomfortable version of the bull case: AMD’s demand might be exactly as strong as guided, and still get rationed by a supply chain that just told everyone it’s running tight.

AMD wasn’t an isolated story that day.

Marvell fell 8%. Intel dropped 7%. Arm had already shed 8% the week before on related worries. The sector lost roughly $200 billion in market value in that one session, reportedly its worst day since June.

That’s not stock-specific risk. That’s the market testing whether the entire AI capex trade still holds together.

This isn’t the first time a supply-chain headline got mistaken for a demand headline this year, and each time, the overreaction has unwound within a few weeks.

And this happens to be the exact week Microsoft and Amazon report earnings, both carrying capex guidance the market will scrutinize line by line.

If either one trims spending plans even slightly, AMD’s server guidance gets questioned by extension, fairly or not.

If both lean in harder on AI infrastructure spending instead, that’s the confidence boost the whole complex needs heading into AMD’s own report two days later.

That’s the real risk the Street’s $600 to $675 targets aren’t pricing in: AMD’s stock doesn’t just trade on AMD anymore.

AMD has been sliding for most of October, and the chart tells the story of a stock that keeps failing every time it tries to bounce.

It’s sitting around $518 now, down from an all-time high of $624.52 set just a month ago, after rebounding off a fresh low near $511 twice in the past week.

The problem is what’s sitting above it. The 20-day average is up near $561, and the 50-day average is all the way up at $570. Every rally into that zone has been sold.

The 14-day relative strength reading is down near 37, which usually signals oversold, not broken, and that’s likely part of why $511 held twice instead of giving way.

If $511 breaks again, the next real floor isn’t far below at $504, a level that’s already acted as support once. Below that, there isn’t much structure until the high $470s, where the lower Bollinger band sits.

For AMD to actually change this story, it needs to reclaim $532 first, then prove it can hold above $560. Until then, every bounce is just a bounce inside a downtrend, not a reversal.

AMD guided server CPU growth above 80% year over year for the back half of this year. On November 3rd, we get the actual number, not the guide. If Data Center growth comes in at the high end of “strong double-digit,” the bull case gets its first real proof point since the September high. If it comes in light, even slightly, the chart’s skepticism gets a lot harder to argue against.

Lam Research pushed some December shipments into March because of component shortages. That’s a timeline problem, not a cancellation, but timelines compound through a supply chain. If Applied Materials or ASML say anything in the next few weeks about tightening further, or similar slippage, that strengthens the case that AMD’s own 2027 ramp could run into real bottlenecks. If the supply picture stabilizes instead, this whole selloff starts looking like an overreaction that unwinds fast. This is the single most underappreciated variable in the whole AMD debate right now.

$532 has acted as resistance on every bounce attempt this month. A close above it, held for more than a day or two, would be the chart finally agreeing with the Street before earnings even arrive to settle anything. Right now the chart hasn’t done that, not once. Watch this level more closely than the headlines between now and November 3rd.

Microsoft reports Wednesday, Amazon reports Thursday, both carrying AI capex guidance the market will pick apart line by line. If either one trims spending plans or softens the language around AI infrastructure investment, AMD’s own guidance gets questioned by extension. If both lean in harder on spending instead, that’s the confidence boost AMD needs heading into its own report two days later.

I covered AMD at its $624 all-time high in AMD Hit $1 Trillion. The Real Number Didn’t Move, asking whether that valuation reflected real ground gained against Nvidia or just sentiment riding high. That question never got answered either way. The stock is now about $100 lower without the debate being settled, and November 3rd is the first real chance since then to get a straight answer. Watch whether management addresses competitive positioning directly, or dodges it again.

I don’t think the Lam Research selloff was really about Lam Research.

I think it was the market finally admitting it’s been pricing AMD on a narrative, not a balance sheet, and any excuse was enough to test that.

Here’s what I keep coming back to: AMD’s own numbers haven’t moved an inch since September. The guidance is the same. The anchor customers are the same. The 2027 doubling target is still on the table.

What’s moved is everyone’s patience for paying $600-plus for a story that still needs two more quarters to prove itself.

I’d rather be early on the stock that proves its numbers than late on the stock that’s already priced like it has.

That’s not a bear case. That’s a repricing of how much the market will pay for AMD’s execution risk, not its growth.

And honestly? I think the Street’s targets end up being roughly right, eventually. Just not before November 3rd forces the first real proof point in a while.

Until this stock reclaims $560 on actual earnings, not sentiment, I’d treat every bounce under that level as a trade, not a conviction call.

AI isn’t a tech trend – it’s a full-blown, multi-trillion dollar race, and 10 companies are already pulling ahead.

These are the innovators driving real revenue, attracting institutional attention, and positioning for massive growth.

Get all 10 tickers in The 10 Best AI Stocks to Own in 2026, free today.

Download Now

The market loves single-cause stories.

A stock drops 8% and everyone wants one explanation: demand died, sell.

Most of the time it’s messier than that. Lam Research’s warning was about equipment, not AI appetite, and it still took AMD down anyway.

That’s the real lesson this week. Prices move on contagion long before they move on facts.

The investors who do well aren’t the ones who ignore the panic. They’re the ones who spend thirty minutes checking whether the panic actually applies to what they own.

Sometimes it does, and this time, it’s still being decided.